Threatscapes for the aeroconservation of birds

Sergio A. Cabrera-Cruz, Scott R. Loss, Jeffrey J. Buler, Emily B. Cohen

The airspace is increasingly cluttered with threats to aerial organisms in the form of anthropogenic structures and vehicles, likely contributing to bird population declines through additive mortality mediated by collisions. In the United States alone, up to and likely exceeding one billion birds die annually due to collisions aloft, and most of these fatalities are related to six threat types: aircraft, buildings, communication towers, power lines, roads, and wind turbines. Collisions happen where birds and threats co-occur; hence, identifying where aerial threats are is a first step in determining avian exposure to risk of collision. Here, we present two collections of high-resolution (30 m) maps that we collectively refer to as threatscapes, illustrating the distribution and magnitude of potential impact for the top six sources of anthropogenic bird collision mortality throughout the contiguous United States. The first collection, the individual-threat maps, contains 285 state-wide spatial rasters, one for each individual aerial threat type present in each state and the District of Columbia, noting that not all threat types were present in every state. The second collection, the cumulative-threat maps, contains 49 spatial rasters displaying the cumulative aerial threat present in each state. To create these threatscapes, we obtained public geospatial data with the distribution of each threat type across the country, and we created risk areas for each individual threat, defined as the two-dimensional spatial footprint where birds may come into contact with structures or vehicles, or where birds may need to take evasive action to avoid colliding with them. We first generated the state-wide individual-threat maps, by assigning to individual risk areas of every state the expected annual bird mortality according to threat-specific published estimates, while also accounting for factors affecting variability of mortality within threat type. Then we created the cumulative-threat maps, by summing the threat-specific maps of each state. Our threatscapes are meant to provide indices of the negative impact that collision threats have on birds and do not represent actual estimates of expected mortality in any given location. We anticipate that these threatscapes will be useful tools for multiple stakeholders to identify priority locations for conservation. For example, one or more individual-threat maps can be combined with bird distribution data to assess avian exposure to risk of collisions with those specific threats, and the cumulative-threat maps can be merged with bird abundance data to identify locations where threats may be contributing to avian population declines. Additionally, both types of threatscapes could be used in combination with geospatial data on widespread sensory pollution agents—for example, artificial light at night, which is linked to bird collisions with several types of structures—to identify areas with interactive effects on birds. Furthermore, these high-resolution threatscapes could facilitate analyzing the impact of anthropogenic obstacles on bird movement, or the response of birds to aerial habitat fragmentation. Overall, these datasets will contribute towards research aimed at identifying drivers of bird population change and toward conservation efforts seeking to slow and reverse the ongoing and worsening decline of North American bird populations.